Median filtering method and device adapting to pipeline architecture and filter

A pipeline and numerical technology, applied in digital adaptive filters, adaptive networks, impedance networks, etc., can solve the problems of strong data dependence and low computing efficiency, and achieve the effect of reducing the amount of calculation, simplifying the logic, and avoiding data structures.

Active Publication Date: 2020-11-17
上海定九康科技股份有限公司
View PDF4 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The embodiment of the present application provides a median filtering method, device and filter adapted to the pipeline architecture, which solves the problem of low calculation efficiency in the prior art due to the strong data dependence before and after operations such as sorting methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Median filtering method and device adapting to pipeline architecture and filter
  • Median filtering method and device adapting to pipeline architecture and filter
  • Median filtering method and device adapting to pipeline architecture and filter

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] refer to figure 1 As shown, this embodiment provides a median filtering method for adapting to the pipeline architecture comprising the steps.

[0046] S101: Receive a set of numeric sequences to be filtered.

[0047] In one embodiment, when signal processing is performed, the analog signal to be processed in the transport layer is received, and after converting an analog signal into a digital signal, a set of numerical sequences is formed. In another embodiment, a set of digital signals is directly accepted. In this embodiment, when filtering is performed, a set of numerical sequences to be processed are directly received.

[0048] S102: The filter window of the preset size is N, acquires the value of the corresponding number in the numerical sequence, forming a first numerical subsequence X 0 , X 1 , ..., x n-1 And from the first numerical subsequence, find the numerical median Miode M 0 .

[0049] This step S102 is a filter initialization operation of this algorithm.

[...

Embodiment 2

[0089] refer to Figure 5 , The present embodiment provides an adaptive median filter device pipelined architecture, the median filtering method is suitable as a pipelined architecture of the first embodiment, the apparatus comprising: a receiving module 101, initialization module 102, an obtaining module 103, processing module 104, circulation module 105, output module 105.

[0090] The receiving module 101 is configured to receive a sequence of a set of values ​​to be filtered.

[0091] Initialization module 102 is configured to obtain the value from the value corresponding to the sequence number according to a preset filter window size of n, x is formed a first numerical sequence 0 , X 1 , ..., x n-1 And the first value from the sequence, find the median value of these values ​​in m 0 .

[0092] Obtaining module 103 is configured in accordance with the value of aligned sequences, the sliding window filter, move it into a value of x n And a value of x removal 0 After the filterin...

Embodiment 3

[0100] The present embodiment provides a filter including a median filter means adapted pipeline architecture as described in Example 2.

[0101] In one application example, the filter median filtering algorithm implemented in the embodiment according to the present embodiment, a floating point (float) on precision data type TI's TMS320C67XX series DSP chip. The DSP chip is a single-core processor, comprising at least an acquisition unit, a comparison unit, calculation unit, wherein the acquisition unit acquires the value, the comparison unit compares the floating point operation, addition and subtraction arithmetic operations calculation unit.

[0102] In this embodiment, by implemented on the chip, and after the code optimization, to achieve the performance per cycle requires only a clk average single filtering performed n / 2 cycles. After numerous experiments, n = 152 for the length of the filter, for example a filtering operation of this embodiment requires only an average of...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a median filtering method and device adapting to a pipeline architecture and a filter. The median filtering method comprises the following steps: S101, receiving a numerical sequence; S102, forming a first numerical value subsequence according to the filtering window, and finding out a numerical value median in the first numerical value subsequence; S103, arranging according to the numerical value sequence, sliding the filtering window, and changing the first numerical value sub-sequence in the first sequence window into a second numerical value sub-sequence after the filtering window is moved into a numerical value and moved out of the numerical value; S104, comparing the shift-in numerical value and the shift-out numerical value with a numerical value median in the first numerical value sub-sequence, and performing logical operation according to a comparison result to obtain a numerical value median in the second numerical value sub-sequence; S105, sliding thefiltering window, and circulating the steps S103 to S104 until the filtering of the numerical sequence is finished; and S106, sequentially storing the numerical value medians, and outputting a numerical value median sequence. According to the filtering algorithm, the logic structure is simple, maintenance of a complex data structure is avoided, the dependence on data in a cycle body is low, pipelining and parallelization are facilitated, and the effect of improving rapid calculation is achieved.

Description

Technical field [0001] The present invention relates to the field of data filtering processing, and more particularly to an intermediate value filtering method, apparatus, and filter of an adaptive pipeline architecture. Background technique [0002] Medium-value filtering is a nonlinear signal processing technology that can effectively suppress noise based on sorting statistical theory, and the basic principle of median filtering is to use this point in numerical sequence (one-dimensional) or digital image (two-dimensional). Medium value of each point value in a neighborhood. [0003] The median filtering algorithm is widely used in signal processing, image processing, automatic control, etc., for the extraction of the signal baseline or eliminates the isolated noise point. For example, in an electrocardiographic signal processing, the baseline is often extracted using the median filtering algorithm, and then the original signal is subtracted from the baseline, which is called "...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): H03H21/00H03H17/04H03H17/02
CPCH03H17/02H03H17/0202H03H17/0211H03H17/0223H03H17/04H03H21/0012H03H21/0029H03H21/0043H03H2017/0207H03H2017/0298H03H2021/007H03H2021/0094
Inventor 徐峰王利明李自明邵猛
Owner 上海定九康科技股份有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products